Government IT Procurement

John Pugh Excerpts

Wednesday 12th January 2011 Westminster Hall

[John Pugh (Southport) (LD)]

It is a pleasure to serve under your chairmanship, Mrs Brooke. Few things are as beguiling as an anoraky discussion about IT on a wet Wednesday afternoon; I am sure that we are all looking forward to it.

I want to add to an improving narrative about the Government and IT, which are not normally considered a good combination; in fact, some would argue that a successful Government IT project is some sort of oxymoron. None the less, the Government are committed to saving money by advancing IT applications. They must save money, and that is one sensible way, on the face of it, to do so. Developing better applications, better use and better value in IT is part of the efficiency agenda that the Government are destined to follow.

That might seem in some ways to be a triumph of hope over experience, but IT is recommended primarily because it offers hope. It offers the hope of speeding up processes or eliminating manual processes, and the opportunity to reduce costs, especially—and, perhaps, regrettably—manpower costs. Those are real benefits in the long term. Unfortunately, the procurer—in this case, the Government—must shell out quite a lot of money in the short term. They must pay up and hope that they get results, and we are all aware of the disasters that occur when they do not.

I encountered a number of those disasters during the previous Parliament through my work on the Select Committee on Public Accounts. Nearly every Department can list one. I do not want to embarrass particular Departments, but I will cite a few. The Lorenzo patient administration software, which I saw demonstrated over the road in Richmond house, has not been implemented anywhere so far, although it was developed at appreciable cost. The Libra project, which affected the Ministry of Justice and the courts, went from £146 million to £232 million. A Department for Transport project ended up unexpectedly spitting out messages in German. The Ministry of Defence, which excels everybody in this respect, developed an infrastructure system that went from £2.3 billion to about £7 billion. I could add other examples with which we are all familiar through our casework: the Child Support Agency, tax credits, the Rural Payments Agency and so on.

I am extremely grateful for some of those illustrations, which were provided to me by the hon. Member for South Norfolk (Mr Bacon), who is something of an expert in the field and who is currently producing a book on why Government IT projects fail. I am sure that it will be available in all good bookshops and via Amazon. From what I have seen of it, it will be a forensic and rattling good read.

However, even the hon. Member for South Norfolk, with his expertise, accepts that there are mitigating factors. Even in the private sector, IT projects are not guaranteed to work. Sainsbury’s had to write off a scheme developed at the cost of £260 million because its failure to operate properly was affecting the company’s share price. The London stock exchange had a heck of a problem over many years developing some of its IT systems, presumably to the detriment of many traders.

Against that, one could say that there are many instances in which the Government are successful, but they are never noticed because they are never of any interest to the press. “Another successful Government IT project accomplished” will never be a headline anywhere. However, I can think of things that we use daily that work satisfactorily that were developed by the Government or Government agencies. I use an Oyster card regularly, for instance. The congestion charge is extraordinarily efficient; sometimes, regrettably, too efficient as far as some of us are concerned. The passport office’s renewal service is good, and the Driver and Vehicle Licensing Agency system for renewing car tax is exemplary as a development of Government via the internet.

There are quiet signs—the Minister will wish to draw attention to them—that the Government have improved incrementally, and maybe substantially, even in the short time that they have been the Government. I believe that it is now common practice to deal with major suppliers, at any rate, as a Government, rather than allowing Departments to be picked off one by one by clever salesmen from software companies and the big IT firms. That is a laudable development, and one modelled on what happens in private enterprise. There are also various hurdles to get over. If some permanent secretary in some Department outside the Cabinet Office thinks that he has an IT project that will run, work and do the trick, he must now satisfy the Cabinet Office that that is actually the case. Such vetting and approval is surely a development that we want to encourage.

Getting things right is a hard art to acquire, and we as a Government and a country might not quite be there. The net effect is that some of the softer savings that we seek—those that are not publicly contentious and do not involve cutting services—will not be secured unless we drill down hard into that area and anatomise what goes wrong.

There are two schools of thought about what goes wrong. One school blames Government procurement: the specification fails, the tendering process is inadequate or civil servants are being suckered by streetwise corporates or conned by the consultants whom they often rope in to advise them. Essentially, they are being sold a pup and getting the wrong thing out of the box. The other school of thought—there is a lot to it—is that it is not the procurement and tendering that are going wrong so much as project management by Government. People talk about deficiencies in client-side expertise in managing, developing and evolving IT projects and about the Government’s relatively poor understanding of software development, saying that there are few experts in the civil service with a thorough grasp of it.

Over the years, there has certainly been appreciable evidence of inadequate cost control and checking, as well as evidence that during the course of a project, Ministers or civil servants often suggest some innovation that ends up costing far more than expected. A classic and highly publicised instance was when the Government asked Boeing to improve the avionics of helicopters. The net effect was a price tag that they could not bear at the time, so we ended up with grounded helicopters, largely because we did not understand what we were asking Boeing for and because the company was certainly not prepared to be thoroughly transparent about what costs might be incurred. Strangely, on that occasion, the MOD did not have the money.

I can think of other instances. EDS was almost taken to court by the Government for the failure of the tax credit system, but part of the problem was obviously the previous Government’s haste to get it up and running, even though EDS genuinely mentioned the problems that it could see. Despite the Government’s bluster about their intention to take EDS to court, they never actually did, ending in a settlement that was more favourable to EDS than some people might have expected given the debacle. That leads one to believe that EDS was not solely to blame for how the project evolved. There is also evidence that in the process of developing the system for the Rural Payments Agency, in particular, civil servants—perhaps the people one expects to monitor the process—took their eye off the ball.

We are therefore torn between two different diagnoses of the ill. It could be procurement; it could be project management. I suggest that they are two sides of the same coin. It might be genuinely insufficiently appreciated how different IT procurement is from other sorts of procurement with which Government are familiar, such as office furniture, tarmac and trains, although even trains have a software dimension nowadays. There is a failure to appreciate that what the Government are ultimately buying is not just things but new ways of doing things. I genuinely think that it is easy for Government to get the procurement of things almost right. I say “almost right”; we MPs are familiar with the printers procured for us by the House. I dare say the House got a good price for the printers, but replacement cartridges cost way over the odds. I must say that that does not seem to be an excellent form of procurement. Equally, there are countless incidents of Government simply not acknowledging that they will have to pay huge amounts in licence fees as part and parcel of what they are ordering, and that often leads to new deals with the companies when Government are alerted to the fact and they simply end up bearing the costs.

With regard to IT hardware procurement, the Public Accounts Committee and the National Audit Office recommended some time ago that it would be sensible for all Government Departments to refresh their hardware every three years and that to leave it longer would be detrimental to true value. In a sense, even that advice is dated. I still work daily with a Power Mac that I ordered in 2001, so its lifespan is already three times longer than the NAO recommends. Therefore, I am not convinced that there is not an art to ordering hardware successfully, but I think that the chances of Government getting that bit right are higher than their chances of getting the new work processes and software processes right.

There are certain implications of Government buying new ways of doing things, rather than simply buying things. To raise a negative point, and one that is not implied, the solution that is often advocated is that we need ever more detailed and expensive tendering processes and more consultant advice. We certainly do not want sloppy tendering or a lack of clarity on objectives, but the problem will not be resolved by having greater detail in the specifications. That is not the answer, unless the name of the game is simply to mind one’s back and prove that one has covered all the bases.

Prolonged over-specification does not seem to be the route out of the problem. Why is that? One thing that it does is put off small suppliers that cannot bear the cost or time constraints of a long tendering process and so cannot stay the pace. One consoling point about the demise of Building Schools for the Future is that its tendering process often excluded many capable and competent people and small and medium-sized enterprises. They might have had a real future role, but they were simply excluded by the nature of the tendering process because they could not match certain preordained criteria.

Another unhelpful consequence of over-specification is that it discourages innovative approaches. I spoke recently with representatives of a computer company that does a lot of business with Government, and they said that they are sometimes hampered not only by the fact that they have a well-defined tendering process, but by the OJEC—Official Journal of the European Community—regulations. They said that Government sometimes interpret those regulations as meaning that they are obliged to tell a company’s competitors if it suggests a new way that it can do what the Government want it to do that is different from what was set out in the original tender, and that is a distinct commercial disadvantage that companies would not welcome.

Over-specification also often leads to the development of complex and expensive bespoke solutions to problems that could otherwise be solved by adapting off-the-peg solutions. I do not want to mention it because it is a sad business, but the Independent Parliamentary Standards Authority had to develop a new expenses system for MPs even though there are probably many available commercially that would have been adequate. I will give another example in which we are perhaps guilty of reinventing the wheel in order to protect our own interests. If a Member wants to look up what they have said in a debate, they can go to Hansard’s search engine, which presumably was specially commissioned, where they will find a rather dismal process. If I try in a week’s time to find what I said about IT, I will probably receive a message telling me that there are no matching criteria, or something like that. Instead, I will go to a commercially-developed website, TheyWorkForYou, which will take me straight to the Hansard pages that the Hansard search engine would not bring me to.

There is much to be said for Government simply making themselves open to a sensible counter offer when there is something out there that meets their needs, rather than demanding bespoke alternatives at every point and turn, because that plays to the big players, and we see evidence of that in Connecting for Health. Highly detailed criteria and inflexibility in the tendering process also rules out the kind of procurement, which works well in the private sector, whereby companies are approached not by individual big companies, but by consortiums that have a variety of skills and can operate in a fluid market.

Another route that we do not want or need to go down in order to find new ways of doing things successfully, which is what software procurement is really about, is having a new army of software programmers and geeks employed by the civil service. Essentially, we need better project managers who can understand the end-user experience, iron out unnecessary complexity and comprehend the implications of changes that are demanded. We need project managers who understand not only the technical implications of project change, but its human aspects. The classic example of failure to do so is the tax credit system, which was developed by a firm that had no familiarity with the circumstances and life conditions of the people who would be applying for tax credits.

I think that there is a route through, but it is not a command and control model. I suggest—I am sure that the Minister is inclined to agree with me—that any successful negotiation of IT and software matters must be a type of conversation, with an acceptance that product delivery requires refinement, hand holding and long-term maintenance. Therefore, we need to get smarter and recognise that it is a specific type of process.

There is a remaining issue that I will try to address in the minutes I have left—I want to leave the Minister 10 minutes in which to respond to the debate. Even if we accept that there is a new way of doing things, that the Cabinet Office will wise up to it and that computer procurement is more like a conversation than an adversarial game, we must recognise that sometimes even conversations turn sour and we end up with ineffectual and underperforming suppliers that are overly costly. What then can Government do? The normal response would be to think that such suppliers should be shown the door, notwithstanding the complication that might ensue for maintenance and further development.

What—if anything—tends to happen depends to some extent on Government’s command over the whole process. If they are looking at a very proprietary solution, such as some quirky software over which a firm has sole and patented control, the options are dramatically reduced. Government can start the project again, which would be expensive and embarrassing, or they can continue to pay up and hope that the costs do not rack up too high, or they can just abandon a particular project and hope that everyone quietly forgets about it. There are incidents of that happening, such as Fujitsu’s withdrawal from Connecting for Health. I suggest that it is in Government’s interests, in most of their procurement, to insist on open standards, although not necessarily open-source software. As a powerful customer, that is exactly what they need to do.

I will make a brief comment on open source software, which is something of an enthusiasm of mine and had the Chancellor’s backing back in 2009. I think that there are real merits in overcoming the Government’s aversion to open source and recognising that it has advanced, that it is producing some very sophisticated SMEs and that it is not comprised entirely of bearded men in pullovers working in garages. There is a widespread belief that it represents a substantial challenge to some proprietary brands of software, and many of the big companies try to stifle open-source encroachment in the public sector.

In 2009, the Chancellor said that open standards, if not open source, could engineer a good deal of saving for the Government. He stated:

“We need to move in the direction of what are known as ‘open standards’—in effect, creating a common language for government IT. This technical change is crucial because it allows different types of software and systems to work side by side in government. At a stroke it means big projects can be split into smaller elements, which can be delivered by different suppliers and then bolted together.”

The article I am reading from is titled,

“When it comes to IT, big is not beautiful”.

It states at the top:

“Agile, modern technology can transform public services and relieve taxpayers of bloated budgets”.

I am sure that we would support that.

There is an overwhelming case for open standards, if not open source, as the Chancellor suggests. If we despair of a construction firm that is not building things properly, we can get in another firm that understands bricks. In mission-critical IT, projects need to be similarly resilient. If that does not mean open source, it certainly means open standards.

To conclude, the Government are moving in the right direction. They are trying to get the boring bit of government right, and that is important—we all subscribe to that—and the Cabinet Office has tried hard to embrace and spread good practice. I would like to hear from the Minister that that will be pushed further and faster, and with more success, because, if we are to believe the Chancellor, the savings would be substantial.

[The Parliamentary Secretary, Cabinet Office (Mr Nick Hurd)]

I welcome you to the Chair, Mrs Brooke. This is the first time I have served under your chairmanship, and it is a real pleasure. I do not know whether you count yourself as one of life’s anoraks. I do not, but I thoroughly enjoyed that speech by the hon. Member for Southport (John Pugh). I congratulate him on securing this debate and, if I may say so, on the elegant way in which he plugged the forthcoming book by my hon. Friend the Member for South Norfolk (Mr Bacon), with whom he served on the Public Accounts Committee.

It is one of the oddities of this place that the Minister for Civil Society should find himself responding to a debate on cost reduction in information and communications technology procurement, but that is my fate as the junior Minister in the Cabinet Office. I should start with an apology to the hon. Member for Southport: this is not my area of day-to-day responsibility, so I am a fish out of water here, but I shall flap to the best of my abilities for the next eight minutes or so. I take this opportunity to reassure him that the Cabinet Office takes seriously the opportunity he has identified to find substantial savings and efficiencies, and better ways for the Government to commission, procure and manage ICT. I shall focus more on procurement, because that is where the Cabinet Office is in a leadership role and where it is particularly focused at present.

The hon. Gentleman was clear about the context in which this debate takes place, including the overriding priority to control the deficit. He will know from his experience on the Public Accounts Committee that there clearly is considerable scope to make savings on ICT, and we are focusing on that. Specifically, there has been a review of major programmes, the introduction of the ICT moratorium and a new, rigorous approval process for procuring consultants and contractors. We expect that up to £1 billion might be saved through the moratoriums in this financial year, if current buying behaviour continues.

In addition to those short-term measures, the Government are committed to delivering sustainable reductions in ICT expenditure—both the cost of goods and services, and the cost of procurement—through the following key initiatives, which are being incorporated. The first one involves managing the Government’s strategic suppliers on a consolidated basis from the centre as the Crown, not from individual Departments—an opportunity that the hon. Gentleman identified. We expect that work to deliver up to £800 million of savings through renegotiating and de-scoping existing contracts with the Government’s top suppliers.

The second initiative involves centralising the procurement of commonly used goods and services, including ICT commodities, to standardise specifications and aggregate the Government’s expenditure more effectively than in the past. That will deliver significant sustainable cost reductions from the existing baseline of £13 billion over the next four years and, in doing so, build on the recommendations made in the Green review.

The third initiative involves streamlining the current process used to source from and contract with suppliers through the application of industry best practice—in this case, “lean” techniques, which have identified the potential to reduce procurement and sourcing project times from an astonishing 85 weeks to 27 weeks, depending on the size, complexity and risk of the procurement, while ensuring compliance with European Union directives.

The fourth initiative involves introducing new spending controls, including a moratorium on any new ICT spend with a value of £1 million or more, and a review of all major projects, a number of which are IT-based. So far, we have reviewed more than 300 Government ICT projects worth nearly £3 billion, and have recommended that more than £1 billion of expenditure be stopped or curtailed. We estimate that the review of major projects has saved £400 million this year. Those are all serious numbers.

We are also developing a strategy that increases the agility of public sector ICT in supporting policy delivery, that achieves greater value for money, that enables effective and efficient digital services and that empowers the big society through greater collaboration and transparency, which is a theme the hon. Gentleman discussed.

Let me elaborate further on each of the initiatives. First, on managing the Government’s strategic suppliers as the Crown rather than as individual Departments, one of the first actions that we took as a new Government was to announce saving targets of £6.2 billion for this financial year. A key part of that involves working with our leading suppliers, including ICT suppliers, to reform our commercial relationships and try to eliminate waste while protecting the fundamental services that we need to provide.

We are driving down the cost of Government IT contracts by negotiating directly with suppliers to achieve cost reductions and efficiency savings in our contracts. As I said, those renegotiation activities are expected to yield £800 million of in-year savings and substantial future savings. In addition, we are commencing reviews of large contracts with more than £100 million remaining contract value to identify further potential savings opportunities.

Secondly, on centralisation of the procurement of commonly used goods and services, we know that the Government buy the same commodity many times but at different price points. We are working to eradicate duplication, apply common specifications, and leverage the Government’s considerable collective buying power more effectively. We are creating a new centralised procurement operating model, and that capability will be mandated across all Government Departments. That transformation of how the Government procure common goods and services through centralised category management is targeted to deliver significant sustainable cost reductions in the region of 25% from the existing baseline of £13 billion.

For each category, there will be a single supply strategy that will deliver a centralised sourcing solution, so that central Government can significantly reduce their spend through aggregation, standardisation and rationalisation. Centralised category management teams will negotiate best-value contracts for central Government Departments based on volume-bound agreements.

[John Pugh]

The Minister mentioned a moratorium. Clearly, a moratorium by itself stops the spend and in doing so saves money, but advocates and supporters of a project who want it to go forward will say that, if it goes forward, it will save far more money than will be saved by virtue of the moratorium. How does the Cabinet Office in its approval mechanism make a judgment about a specific project? Does it take a broad view on how credible a project is or how secure the savings are, or is it simply the case that a certain number of projects need to be stopped in order to meet a certain figure?

[Mr Hurd]

It is more the case that the moratorium provides an opportunity to go through the discipline of looking closely at what is actually going on.

In my last minute, I would like to expand a little on the third pillar of our approach, which involves improving procurement processes. We know that bid costs range from £20,000 to £200,000 for every month. We are determined to reduce that, and the 6,000 pages of guidance on procurement within Government. The “lean” study has set out the task ahead of us: to refine and introduce a lean, new procurement process.

The point of all that is to try to reassure the hon. Gentleman that the Government take extremely seriously the need and the opportunity to be more efficient in how they procure IT and manage it on an ongoing basis.